Professional Documents
Culture Documents
ebook download (eBook PDF) Business Statistics A First Course First Canadian Edition all chapter
ebook download (eBook PDF) Business Statistics A First Course First Canadian Edition all chapter
http://ebooksecure.com/product/original-pdf-business-statistics-
a-first-course-second-2nd-canadian-edition/
http://ebooksecure.com/product/ebook-pdf-business-statistics-a-
first-course-7th-edition/
http://ebooksecure.com/product/ebook-pdf-business-statistics-a-
first-course-8th-edition/
http://ebooksecure.com/product/ebook-pdf-business-statistics-a-
first-course-3rd-edition/
Business Statistics: A First Course 7th Edition (eBook
PDF)
http://ebooksecure.com/product/business-statistics-a-first-
course-7th-edition-ebook-pdf/
http://ebooksecure.com/product/ebook-pdf-business-statistics-a-
first-course-6th-edition/
http://ebooksecure.com/product/ebook-pdf-business-statistics-a-
first-course-global-edition-8th-edition/
http://ebooksecure.com/product/ebook-pdf-first-course-in-
statistics-a-11th-edition/
http://ebooksecure.com/product/ebook-pdf-business-statistics-a-
first-course-3rd-edition-by-norean-d-sharpe/
Contents vii
Appendixes
A Answers A-1
B Tables and Selected Formulas A-31
8
CHAPTER
Random Variables and
Probability Models
C O N N EC T I O N S: C H A P T E R
With the ideas of probability we discussed in Chapter 7 we now
have the tools to create specific models that describe where the
data came from. In this chapter we will learn about two types of
models—discrete and continuous—that correspond to the two
types of data we have worked with—categorical and quantitative.
The models will be used all the way through the rest of the text;
they are the key to understanding and interpreting data. This is
also where the famous bell-shaped curve will be introduced.
Manulife Financial
W
hat company could be more Canadian than
one whose first president was the first Prime
Minister of Canada, Sir John A. Macdonald?
In-Text Examples
LEARNING OBJECTIVES The Manufacturers Life Insurance Company Real business examples motivate the
was founded in 1887, and a mere ten years later expanded its
❶ Find the expected value and standard operations into China and Hong Kong. Manulife now operates discussions, often returning to the
deviation of a random variable or in 21 countries, most notably in the United States, through its chapter-opening company.
combination of random variables. subsidiary, John Hancock Insurance.
Manulife Financial is the largest insurance company in Canada,
❷ Use the Binomial and Geometric
the second largest in the North America (after MetLife, well-
models to find probabilities.
known because of their use of Snoopy, the dog from the cartoon
❸ Use the 68–95–99.7 Rule to find 99.9%?
strip “Peanuts”), and the world’s fifth largest, based on market 4.3 Charts
probabilities or cutoff values. If you are careful to add the percentages in Table 4.2,
capitalization. At the end of 2011, its total assets
youexceeded US total is 99.99%. Of course the real
will notice the The Area Principle
❹ Use the Normal model to find prob- $450 billion and total employees surpassed 25total 000.hasManulife
to be 100.00%.
is alsoThe discrepancy is due to indi-
Now that we have a frequency table, we’re ready to fol-
abilities, cutoff values, and the mean a financial services provider and holds Manulife Bank of Canadabeing
vidual percentages as rounded. You’ll often see this in low the three rules of data analysis and make a picture of
tables of percents, sometimes with explanatory footnotes.
and standard deviation. a wholly-owned subsidiary. the data. But we can’t make just any picture; a bad pic-
Manulife changed from a joint stock company to a mutual ture can distort our understanding rather than help it.
❺ Compute and use z-scores for For example, here’s a graph of the frequencies of Table
comparisons. company, privately owned by its policyholders in 1958. It was
4.1. What impression do you get of the relative frequen-
demutualized in 1999 and shares of the holding company Manulife cies of visits from each province?
❻ Use the Normal model to approxi- Financial Corporation began trading on the TorontoIt is often also very useful to rearrange the order of the
Stock The figure does not accurately represent the infor-
mate the Binomial and compute Exchange (TSX), New York Stock Exchange (NYSE), categories, and the bars, so that they go from highest to
and others. mation in the table. What’s gone wrong? The height of
probabilities. lowest (or lowest to highest). This version is referred to the images in the figure do match the percentages in the
In 2002, China gave approval to Manulife-Sinochem Life but it is just a variation of the bar chart
as a Pareto chart, table. But our eyes tend to be more impressed by the
we have discussed.
Insurance Co. Ltd. to open a branch office in Guangzhou, China,Here is another look at Figure 4.3, area (or perhaps even the volume) than by other aspects
with reordered, and horizontal, bars.
of each image, and it’s that aspect of the image that we
notice.
Nova Scotia 209 Since there were nearly three times as many vis-
Manitoba
its from Ontario and BC as from Quebec, the images
depicting the number from Ontario and from BC is
almost three times higher than the image for Que-
Province
Alberta
bec, but it occupies almost nine times the area, since
Quebec both the height and the width were increased three-
M08_SHAR4837_01_SE_C08.indd 209 5/15/14 2:21 PM
BC fold to keep the image looking proportional. As you
can see from the frequency table, that isn’t a correct
Ontario impression.
The best data displays observe a fundamental prin-
0% 10% 20% 30% 40% ciple of graphing data called the area principle, which
Percent of Visits says that the area occupied by a part of the graph should
correspond to the magnitude of the value it represents.
x
…To the Boardroom
We lead students through the process of making business decisions with data. The first step is planning how to tackle a
problem, the second is doing the calculations, and the third is reporting the results and conclusions. In each chapter, we
apply the new concepts learned in Guided Example. Examples are structured to the way statisticians approach and solve
problems. These step-by-step examples show students how to produce the kind of solutions and reports that clients expect
to see.
GUIDED EXAMPLE
GUIDED EXAMPLE NewYork
New YorkStock
StockExchange
Exchange Trading
Trading Volume
Volume
statistics. This is what most people think Mechanics Plot the side-by-side boxplots
3.0
Daily Volume (Billions)
Statistics is about. But the computa- of the data. 2.7 *
2.4
tions don’t tell the whole story. 2.1
1.8
M05_SHAR4837_01_SE_C05.indd 102 1.5 5/6/14 4:39 PM
1.2
0.9
0.6 *
REPORT what you’ve learned. Until 1 2 3 4 5 6 7 8 9 10 11 12
Month
you’ve explained your results in the
context of the business question in your
Plan, the job isn’t done. We present the Conclusion Report what you’ve learned MEMO:
about the data and any recommended action
report step as a memo to emphasize the or analysis. Re: Research on trading volume of the NYSE
decision aspect of each example. We have examined the daily sales volume on the
NYSE (number of shares traded) for each month
of 2006. As the attached display shows, sales
volume follows a seasonal pattern with lower volume
in March and August. The highest median trading
activity is found in November. The variability of trad-
ing volume also shows a pattern. June and December
have higher variability than the rest, and March
has noticeably less variability. There were several
unusually high-volume days that bear investigation
and one extremely low-volume day in November.
xi
Promoting Understanding
What Can Go Wrong?
The most common mistakes for those WHAT CAN GO WRONG?
new to statistical analysis usually involve ● Don’t violate the area principle. This is probably the most common mistake
in a graphical display. Violations of the area principle are often made for the
misusing a method, not miscalculating a sake of artistic presentation. Earlier we quoted Tufte, who said that “the only
worse design than a pie chart is several of them...” Here is another take. The
statistic. We acknowledge these mistakes only worse design than a 2-D pie chart is a 3-D pie chart. If you insist on using
with What Can Go Wrong?, found at the a pie chart, don’t compound the visual perception difficulty by adding irrel-
evant and misleading perspective. While we’re on the subject, don’t use 3-D
end of each chapter. Our goal is to arm bar chart either (see below). Here, for example, are two versions of the same
pie chart for the Regional Preference data.
students with the tools to detect statisti-
cal errors and offer practice in debunking Neither Disagree
Agree
Completely Neither Disagree
Agree
Completely
misuses of statistics. Nor Agree Nor Agree
Math Box The one on the left looks interesting, doesn’t it? But showing the pie three
dimensionally on a slant violates the area principle and makes it much more
The mathematical underpin- difficult to compare fractions of the whole made up of each category of
the response—the principal feature that a pie chart ought to show. Those
nings of statistical methods and sections of the pie in the forefront are emphasized. If you want to mislead the
viewer, put the “bad news” category in the back and the “good news” category
concepts are set apart to avoid in the front. The viewer sees not only the top of the pie but the side as well.
interrupting the explanation of ● Keep it honest. Here’s
Use a pie chart that displays data on the percentage of high
Marijuana
school students who engage in specified dangerous behaviours. What’s wrong
the topic at hand. We use these with this plot?
26.7%
50.0%
Use
Alcohol
derivations to increase students’ 31.5%
course, showing the pie on a slant makes it even harder to detect the error.
# of Airline Passengers Searched
3000
MATH BOX
2500
We want to find the mean (expected value) of random variable X, using a geometric method
with probability of success p. 2000
1500
First write the x 1 2 3 4 ...
probabilities: 1000
P(X 5 x) p qp q2p q3p ...
500
The expected value is: E1X2 = 1p + 2qp + 3q2p + 4q3p + g
0
2 3 77–78 79–80 81–82 83–84 85–86 87–88 89–90 91–92 93–94 95–96 1997 1998 1999
Let p = 1 - q: = 11 - q2 + 2q11 - q2 + 3q 11 - q2 + 4q 11 - q2 + g
2 2 3 3 4 Year
Simplify: = 1 - q + 2q - 2q + 3q - 3q + 4q - 4q + g
That’s an infinite = 1 + q + q2 + q3 + g Finding the correlation coefficient by hand
Looks like things didn’t
To find change much
the correlation in thebyfinal
coefficient hand,years of the twentieth
geometric series, with 1
first term 1 and = century—untilwe’ll
you use
reada formula
the bar labels and see
in original that
units, the last
rather thanthree bars represent
1-q single years, while all the
z-scores. others
This are for
will save us the of years.
work
pairs Thetofalse depth makes it
of having
common ratio q:
So, finally
1
E1X2 = . hstandardize
d each
h individual
bl data value first. Start with
p the summary statistics for both variables: x, y, sx, and
sy. Then find the deviations as we did for the standard
deviation, but now in both x and y: (x - x) and (y - y)
. For each data pair, multiply these deviations together:
(x - x) * (y - y). Add the products up for all data pairs.
Finally, divide the sum by the product of (n - 1) * sx * sy
to get the correlation coefficient.
Here we go.
x 6 10 14 19 21
Suppose the data pairs are:
y 5 3 7 8 12
Deviations Deviations
By Hand in x
6 - 14 = - 8
in y
5 - 7 = -2
Product
- 8 * - 2 = 16
explain formulas and help students through the calcu- Finally, we divide by (n - 1) * sx * sy = (5 - 1) * 6.20 *
3.39 = 84.07
lation of a worked example. The ratio is the correlation coefficient:
r = 72>84.07 = 0.856
xii
M04_SHAR4837_01_SE_C04.indd 71 5/1/14 5:05 PM
Integrating Technology
Technology Help
In business, Statistics is prac- TECHNOLOGY HELP: Displaying and Summarizing Quantitative Variables
tised with computers. We Almost any program that displays data can make a histogram, but It is usually easy to read the results and identify each com-
offer specific guidance for some will do a better job of determining where the bars should puted summary. You should be able to read the summary statistics
start and how they should partition the span of the data (see the produced by any computer package.
Excel®, often with an anno- art on the next page). Packages often provide many more summary statistics than
tated example to help stu- Many statistics packages offer a prepackaged collection of
summary measures. The result might look like this:
you need. Of course, some of these may not be appropriate when
the data are skewed or have outliers. It is your responsibility to
dents get started with the Variable: Weight
check a histogram or stem-and-leaf display and decide which
N = 234 summary statistics to use.
technology of their choice. Mean = 143.3 Median = 139 It is common for packages to report summary statistics to
St. Dev = 11.1 IQR = 14 many decimal places of “accuracy.” Of course, it is rare to find data
Alternatively, a package might make a table for several that have such accuracy in the original measurements. The ability
variables and summary measures: to calculate to six or seven digits beyond the decimal point doesn’t
mean that those digits have any meaning. Generally, it’s a good idea
Variable N mean median stdev IQR to round these values, allowing perhaps one more digit of precision
Weight 234 143.3 139 11.1 14 than was given in the original data.
Height 234 68.3 68.1 4.3 5 Displays and summaries of quantitative variables are among
Score 234 86 88 9 5 the simplest things you can do in most statistics packages.
EXCEL
xiii
Checking Understanding
Just Checking
Once or twice per chapter, Just Checking asks JUST CHECKING
students to stop and think about what they’ve So that they can balance their inventory, an optometric practice collects the following data about its patients.
Sex
Females 4 16 12 32
Total 10 36 18 64
students can easily check their work.
1 What percent of females are far-sighted?
2 What percent of near-sighted customers are female?
3 What percent of all customers are far-sighted females?
4 What’s the distribution of Eye Condition?
5 What’s the conditional distribution of Eye Condition for males?
6 Compare the percent who are female among near-sighted customers to the percent of all customers who are female.
7 Does it seem that Eye Condition and Sex might be dependent? Explain.
M04_SHAR4837_01_SE_C04.indd 67
What Have We Learned? 5/1/14 5:04 PM
Skills
things down precisely.
Terms
• Recognize when a variable is categorical and choose an appropriate display for it.
Case A case is an individual about whom or which we have data.
• Understand how to examine the association between categorical variables by
Categorical variable A variable that names categories (whether with words or numerals) is called
comparing conditional and marginal percentages.
categorical.
Context The context ideally tells Who was measured, What was measured, How• theSummarize
data the distribution of a categorical variable with a frequency table.
were collected, Where the data were collected, and When and Why the study was
performed.
• Display the distribution of a categorical variable with a bar chart.
• Construct and examine a contingency table.
Cross-sectional data Data taken from situations that vary over time but measured at a single time
instant are said to be a cross-section of the time series. • Construct and examine displays of the conditional distributions of one variable
for two or more groups.
Data Systematically recorded information, whether numbers or labels, together with its
context.
• Describe the distribution of a categorical variable in terms of its possible values
Data table and relative
An arrangement of data in which each row represents a case and each column frequencies.
represents a variable.
• Describe any anomalies or extraordinary features revealed by the display of a
Experiment variable.
A study in which the researcher manipulates factor levels to assess the effect of the
factor on the response.
• Describe and discuss patterns found in a contingency table and associated dis-
Experimental unit An individual in a study for which or for whom data values are recorded. plays
Humanof conditional distributions.
experimental units are usually called subjects or participants.
Identifier variable A categorical variable that records a unique value for each case, used to name or
identify it.
Nominal variable The term “nominal” can be applied to data whose values are used only to name
categories.
xiv
Tackling Problems
Exercises
We have worked hard to ensure that the exercises contain relevant and modern ques-
tions with real data. The exercises generally start with a straightforward application of
the ideas, then tackle larger problems. Many break a problem into several parts to help
guide the student through the logic of a complete analysis. Finally, there are exercises
that ask students to synthesize and incorporate their ideas with less guidance. Data sets
for exercises marked with T can be found on MyStatLab.
EXERCISES
Homes for Sale for 180 countries as grouped by the World Bank. Each point rep-
6 resents one of the years from 1970 to 2007. The output of a
regression analysis follows. LO➋
5
Price ($000 000)
Countries (%)
2 5
1
4
0
5.0 10.0 15.0 3
Rooms
2
a) Is there an association?
b) Check the assumptions and conditions for correlation.
T 22. WestJet. WestJet is Canada’s second largest Canadian air 1 2 3 4 5 6
carrier, and the ninth-largest in North America by passengers Annual GDP Growth of Developed
Countries (%)
carried (over 17 million in 2012). Founded in 1996, WestJet is a
public company with over 9000 employees, non-unionized, and
not part of any airline alliance. Here are two scatterplots of data Dependent variable: GDP Growth Developing Countries
R2 5 20.81%
from 1998–2012. The first shows the growth in annual revenue
s 5 1.244
over time. The second shows the relationship between annual
revenue and number of passengers (called “segment guests”). Variable Coefficient
Intercept 3.46
Describe the relationships seen in the two scatterplots. Are they GDP Growth Developed 0.433
linear? Are there any unusual features or data points? LO❶ Countries
WestJet Revenue Growth a) Check the assumptions and conditions for the linear model.
4000 b) Explain the meaning of R2 in this context.
3500 c) What are the cases in this model?
3000
M06_SHAR4837_01_SE_C06.indd 168 T 24. European GDP growth. Is economic growth in5/10/14 Europe
1:51 PM
Revenue ($M)
4000
6
3500
3000 5
Revenue ($M)
2500
2000 4
1500
3
1000
500
2
0
0 5000 10 000 15 000 20 000
1
Passengers (000s)
0
ETHICS IN AC TION T 23. GDP growth. Is economic growth in the developing world
related to growth in the industrialized countries? Here’s a
Advocacy groups for equity and diversity in the workplace completed the six-week online diversity
scatterplot of theeducation
growthprogram
(in % of Gross Domestic Product) of –2 0 2 4 6
often cite middle managers as an obstacle to many organ- and was retested. The average on the testcountries
the developing after completing
vs. the growth of developed countries Annual GDP Growth of United States (%)
izations’ efforts to be more inclusive. In response to this the online program was 69.15. Although the group achieved
concern, Michael Schrute, the CEO for a large manufacturing a higher mean test score after completing the program, the
company, asked Albert Fredericks, VP of Human Resources, two-sample t-test revealed that this average test score was
to look into the possibility of instituting some type of divers- not significantly higher than the average prior to completing
the online program (t = - 0.94, P-value = 0.176). Albert was
ity training for the company’s middle managers. One option
under consideration was an online education program that
focused on cultural diversity, gender equity, and disability
not surprised and began to explore more traditional diversity
educational programs.
Ethics in Action
awareness. Although cost-effective, Albert suspected that
an online program would not be as effective as traditional
ETHICAL ISSUE The pretest and posttest design violates Our ethics vignettes in each chapter illustrate some
independence and therefore the two-sample t-test
M06_SHAR4837_01_SE_C06.indd 172 is not appro- 5/10/14 1:51 PM
training for middle managers. In order to evaluate the online
program under consideration, 20 middle managers were
priate (related to Item A, ASA Ethical Guidelines). of the judgments needed in statistical analysis,
selected to participate. Before beginning, they were given a
test to assess their knowledge of and sensitivity to various
ETHICAL SOLUTION Use the correct test. The two-sample
t-test is not appropriate for these data. Use a paired t-test
identify possible errors, link the issues to the ASA’s
diversity and equity issues. Out of a possible perfect score of
100, the test average was 63.65. Each of the 20 managers then
instead. Using the correct test shows that the online diversity
education program was effective.
Ethical Guidelines, and then propose ethically and
statistically sound alternatives.
xv
xvi PREFACE
This book recognizes both the changing curriculum and the changing pedagogy for
teaching introductory Statistics. It focuses on application, streamlines and reorganizes
topics, sheds unneeded theoretical details, and recognizes learning styles of the cur-
rent generation of students, making it an attractive choice for single-term courses at
Canadian business schools.
The Canadian edition uses illustrative case studies and chapter exercises that focus
on business and management of Canadian companies, large and small. The writing
has Canadian style, perspective, and sensibility, not just changes to spelling and met-
ric measurements. The book recognizes that Canada has public policy, governmental
structure and mandate (federal, provincial, and municipal), economic systems (bank-
ing and finance), social services, health care, sports, and entertainment that are differ-
ent from those of the United States. While the book is indeed Canadian, it also shows
how Statistics has no geographical borders and is a vital part of building and enhanc-
ing the global community.
Most chapter-opening examples reflect a Canadian situation that is then used
within the chapter to illustrate new concepts. Throughout the chapters, exercises use
Canadian sources and examples, as do most Mini Case Study Projects.
The Canadian edition has two new features on the chapter opener, to help map
the chapter content: Learning Objectives, which are then tied to the end-of-chapter
exercises, and a Connections box that situates the content within the text as a whole.
New and expanded topics, new sidebars, and added warnings in What Can Go Wrong
features throughout the chapters provide further clarity and context for students.
Chapter 16, Statistical Modelling and the World of Business Statistics, is a brand-
new capstone chapter. It provides both a framework for the inference methods devel-
oped in Chapters 9 through 15 and a method for choosing an appropriate procedure
for a given data analysis situation.
Here are some highlights of what else is new in the Canadian edition:
• Chapter 1 encourages students’ interest in and understanding of the discipline
through the revised chapter opener and a brand-new Section 1.1, The Role of
Statistics in the World. This chapter also includes new material to explain and
expand on the concept of variation.
• Chapter 2 expands Section 2.2, Types of Variables (and Types of Data), to
include topics such as “imaginary” units or underlying measurement scale, the
creation of quantitative variables as summed ordinal scales, and string variables.
The new Section 2.3, Data Quality, categorizes data as “good” versus “bad” and
explores various aspects of data quality.
• Chapter 3 provides additional explanations of population, parameter, sample, sta-
tistic, and sampling frame and sidebar discussions that help contextualize the con-
cept of polls and surveys.
• Chapter 4 introduces alternatives and variations on bar charts—dot plot, Pareto
chart—and segmented bar charts as an alternative to pie charts.
• Chapter 5 has new historical notes on the term histogram, the origins of stem-
and-leaf displays, and the use of x-bar notation; an added explanation of har-
monic and geometric means, with illustrative examples; and the probability
explanation of the n – 1 denominator of sample standard deviation.
• Chapter 6 has a new discussion on regression to the mean, with graphical
illustration.
The Canadian Edition of Business Statistics xvii
PREFACE
W
e set out to write a book for business students that answers the simple
question: “How can I make better decisions?” As entrepreneurs and
consultants, we know that statistics are essential to survive and thrive
in today’s competitive environment. As educators, we’ve seen a discon-
nect between the way Statistics is taught to business students and the way it is used in
making business decisions. In Business Statistics: A First Course, we try to close the gap
between theory and practice by presenting statistical methods so they are both relevant
and interesting to students.
The data that inform a business decision have a story to tell, and the role of statis-
tics is to help us hear that story clearly. Like other textbooks, Business Statistics: A First
Course teaches how to select and perform inferential tests and highlights definitions
and formulas. But, unlike other textbooks, Business Statistics: A First Course also teaches
the “why” and insists that results be reported in the context of business decisions. Stu-
dents will come away knowing how to use statistics to make better business decisions
and how to effectively communicate both their analysis and recommendations.
Business Statistics: A First Course is written with the understanding that today Statis-
tics is practised with technology. This insight informs everything from our choice of
forms for equations to our extensive use of real data. But more important, understand-
ing the value of technology allows us to focus on teaching statistical thinking rather
than calculation. The questions that motivate each of our hundreds of examples are
not “how do you find the answer?” but “how do you think about the answer, and how
does it help you make a better decision?”
Our focus on statistical thinking ties the chapters of the book together. An intro-
ductory business statistics course covers an overwhelming number of new terms, con-
cepts, and methods. But they have a central core: how we can understand more about
the world and make better decisions by understanding what the data tell us. From this
perspective, students can see that the many ways to draw inferences from data are sev-
eral applications of the same core concepts.
and the exercises reinforce this. When we graph data, we are able to see struc-
tures, or patterns, that we can’t see otherwise. These patterns often raise new
questions and guide our statistical analysis and case analysis process. Emphasiz-
ing graphics throughout the book helps students see that the structures we look
for in graphs help illuminate sophisticated concepts.
• Consistency. We work hard to avoid the “do as we say, not as we do” trap. Hav-
ing taught the importance of plotting data and checking assumptions and condi-
tions, we are careful to model that behaviour throughout the book. (Check the
exercises in the chapter on multiple regression and you’ll find us still requir-
ing and demonstrating the plots and checks that were introduced in the early
chapters.) This consistency helps reinforce these fundamental principles.
• The need to read. Students who plan just to skim the book may find that im-
portant concepts, definitions, and sample solutions are not always set aside in
boxes. This book needs to be read, so we’ve tried to make the reading experience
enjoyable.
Coverage
The topics covered in a Business Statistics course are generally consistent and man-
dated by our students’ needs in their studies and in their future professions. But the
order of these topics and the relative emphasis given to each is not well established. In
Business Statistics: A First Course, you may encounter some topics sooner or later than
you expected. Although we have written many chapters specifically so they can be
taught in a different order, we urge you to consider the order we have chosen.
We’ve been guided in the order of topics by the fundamental principle that this
should be a coherent course in which concepts and methods fit together to give stu-
dents a new understanding of how reasoning with data can uncover new and impor-
tant truths. We have tried to ensure that each new topic fits into the growing structure
of understanding that we hope students will build. For example, we teach inference
concepts with proportions first and then with means. Students have a wider expe-
rience with proportions, seeing them in polls and advertising. And by starting with
proportions, we can teach inference with the Normal model and then introduce infer-
ence for means with the Student’s t-distribution.
We introduce the concepts of association, correlation, and regression early in
Business Statistics: A First Course. Our experience in the classroom shows that intro-
ducing students to these fundamental ideas early motivates them at the beginning
of the course. Later in the semester, when we discuss inference, students recall what
they have learned and find it natural and relatively easy to build on the fundamental
concepts they experienced by exploring data with these methods. In an introductory
course, the relative emphasis placed on topics requires planning by the instructor.
Introductory Business Statistics courses often have limited time, so it can be hard to
get to the important topics of multiple regression and model building with enough
time to treat them thoroughly. We’ve been guided in our choice of what to empha-
size by the GAISE (Guidelines for Assessment and Instruction in Statistics Educa-
tion) Report,1 which resulted from extensive experience in how students best learn
statistics. Those recommendations, which received wide support and adoption across
North America (Statistical Society of Canada and American Statistical Association),
urge (among other detailed suggestions) that Statistics education should:
1. Emphasize statistical literacy and develop statistical thinking;
2. Use real data;
3. Stress conceptual understanding rather than mere knowledge of procedures;
1
www.amstat.org/education/gaise
xx PREFACE
Flexible Syllabus. Business Statistics: A First Course follows the GAISE Guidelines.
The committee that developed these guidelines was made up of innovative educa-
tors in Statistics education, including one of the authors of this textbook (Velleman).
Those guidelines call for presenting students with real data early and throughout the
course and emphasizing real-world decisions and understanding as the final step of
any statistical analysis. Following that advice, we have placed an introductory (explor-
atory, non-inference) section on regression early in the text (Chapter 6).
However, we are acutely aware of the challenge that Statistics instructors face in
covering essential material in a limited period of time. For that reason we offer sug-
gested modifications that may better fit your teaching style and requirements:
• Chapter 3 (Surveys and Sampling) can be taught just before Chapter 7 (Random-
ness and Probability), where it would provide motivation for randomness.
• Chapter 6 (Correlation and Linear Regression [without inference]) can be taught
just before Chapter 14 (Inference for Regression), where it would form a larger
lesson on regression analysis through inference and can be followed immediately
by Chapter 15’s discussion of multiple regression.
• Chapter 7 (Randomness and Probability) can be excluded if probability is taught
in a separate course (e.g. quantitative decision-making), or for a more data-
focused course.
• We introduce inference with proportions. We find this works better for stu-
dents (who know about polls and margins of error in that context), and makes
the presentation simpler. If you prefer to introduce inference with means, you
can teach Chapter 11 (Confidence Intervals and Hypothesis Tests for Means)
before Chapter 10 (Testing Hypotheses about Proportions). Note, however, that
we teach inference for means without the assumption that the population stan-
dard deviation is known, and thus emphasize Student’s t. That reflects the almost
universal practice in real business analyses and in all statistics software. It also
frees students of any concern about how big a sample should be before “switch-
ing” from t to Normal—with modern technology, there is no need to switch. And
no statistics package or regression analysis does switch.
• Chapters 12 (Comparing Two Groups) and 13 (Inference for Counts: Chi-Square
Tests) are omitted by some instructors.
• Chapter 15 introduces multiple regression, which can also be omitted.
• Chapter 16 is designed to be a capstone chapter, by summarizing inference in
the framework of statistical modelling. For courses that do not cover regression
models, this chapter is optional.
Features
A textbook isn’t just words on a page: it is many features that come together to form a
big picture. The features in Business Statistics: A First Course provide a real-world con-
text for concepts, help students apply these concepts, promote problem-solving, and
Preface xxi
integrate technology—all of which help students understand and see the big picture
of Business Statistics.
Motivating Examples. Each chapter opens with a motivating example, often taken
from the authors’ consulting experiences. These companies—such as Angus Reid,
Mountain Equipment Co-op, Manulife Financial, and Canada Goose—enhance and
illustrate the story of each chapter and show students how and why statistical thinking
is so vital to modern business decision-making. We analyze the data from those com-
panies throughout the chapter.
Learning Objectives. Each chapter begins with a list of learning objectives. These
LEARNING OBJECTIVES are brief but clear statements about what students are expected to know and be able
to demonstrate by the end of chapter (or the end of the course). These will also help
instructors with course planning and classroom delivery. Each end-of-chapter exer-
cise references one or more learning objectives to guide students’ assessment of their
progress with the material.
Connections. Although the authors (and, we hope, the instructors) know how the
CONNECTIONS chapters of a textbook are related to each other and understand the logic behind
the sequencing, it may not be clear to students until they reach the end of the book.
The Connections boxes explain how the current chapter is related to the previous
chapter(s), and why its position in the sequence is appropriate.
Mini Case Study Projects. Each chapter includes one or two Mini Case Study Proj-
MINI CASE STUDY PROJECTS ects that use real data and ask students to investigate a question or make a decision.
M01_SHAR4837_01_SE_C01.indd 1
Students define the objective, plan the process, complete the analysis, and report a
5/1/14 4:42 PM
conclusion. Data for the Mini Case Study Projects are available on MyStatLab.
What Can Go Wrong? Each chapter contains an innovative section called What Can
Go Wrong? that highlights the most common statistical errors and the misconcep-
tions people have about Statistics. The most common mistakes for the new user of
statistical methods involve misusing a method, not miscalculating a statistic. Most of
the mistakes we discuss have been experienced by the authors in a business context
rather than a classroom situation. One of our goals is to arm students with the tools to
detect statistical errors and to offer practice in recognizing when Statistics is misused,
whether intentional or not. In this spirit, some of our exercises probe the understand-
ing of such errors.
REALITY CHECK Reality Check. We regularly remind students that Statistics is about understanding
the world and making decisions with data. Results that make no sense are probably
xxii PREFACE
wrong, no matter how carefully we think we did the calculations. Mistakes are often
easy to spot with a little thought, so we ask students to stop for a reality check before
interpreting results.
Notation Alert. Throughout this book, we emphasize the importance of clear com-
munication. Proper notation is part of the vocabulary of Statistics, but it can be daunt-
ing. Students who know that in Algebra n can stand for any variable may be surprised
to learn that in Statistics n is always the sample size. Statisticians dedicate many letters
and symbols for specific meanings (b, e, n, p, q, r, s, t, and z, along with many Greek
letters all carry special connotations). Students learn more effectively when they are
clear about the letters and symbols statisticians use.
ETHICS IN ACTION Ethics in Action. Students are often surprised to learn that Statistics is not just plug-
ging numbers into formulas. Most statistical analyses require a fair amount of judg-
ment. The best guidance for these judgments is that we make an honest and ethical
attempt to learn the truth. Anything less than that can lead to poor and even dan-
gerous decisions. The Ethics in Action vignettes in each chapter illustrate some of the
judgments needed in statistical analyses, identify possible errors, link the issues to
widely accepted ethical guidelines in Statistics, and then propose ethically and statisti-
cally sound alternative approaches.
Exercises. We’ve worked hard to ensure that exercises contain relevant, modern,
and real-world questions. Many come from news stories; some come from recent
research articles. Whenever possible, the data are on MyStatLab (always in a variety
of formats) so students can explore them further. Sometimes, because of the size of
the data set, the data are only available electronically. Throughout the book, we pair
the exercises so that each odd-numbered exercise (with answer in the back of the
book) is followed by an even-numbered exercise on the same Statistics topic. Exer-
cises are roughly ordered within each chapter by both topic and by level of difficulty.
Data Sources. Most of the data used in examples and exercises are from real-world
sources, and we list many sources in this edition. Whenever we can, we include
references to the internet data sources used, often in the form of URLs. As internet
users (and thus, our students) know well, URLs can “break” as websites evolve. To
minimize the impact of such changes, we point as high in the address tree as is prac-
tical. Moreover, the data themselves often change as more recent values become
Preface xxiii
available. The data we use are usually on MyStatLab. If you seek the data—or an
updated version of the data—on the internet, we try to direct you to a good starting
point.
TECHNOLOGY HELP Technology Help. In business, Statistics is practised with computers, but not with
a single software platform. Instead of emphasizing a particular statistics program, at
the end of each chapter, we summarize what students can find in the most common
packages, often with annotated output. We then offer specific guidance for one of the
most common packages, Excel®, to help students get started with the software of their
choice.
xxiv PREFACE
Supplements
1.D. The copyright laws of the place where you are located also
govern what you can do with this work. Copyright laws in most
countries are in a constant state of change. If you are outside
the United States, check the laws of your country in addition to
the terms of this agreement before downloading, copying,
displaying, performing, distributing or creating derivative works
based on this work or any other Project Gutenberg™ work. The
Foundation makes no representations concerning the copyright
status of any work in any country other than the United States.
1.E.6. You may convert to and distribute this work in any binary,
compressed, marked up, nonproprietary or proprietary form,
including any word processing or hypertext form. However, if
you provide access to or distribute copies of a Project
Gutenberg™ work in a format other than “Plain Vanilla ASCII” or
other format used in the official version posted on the official
Project Gutenberg™ website (www.gutenberg.org), you must, at
no additional cost, fee or expense to the user, provide a copy, a
means of exporting a copy, or a means of obtaining a copy upon
request, of the work in its original “Plain Vanilla ASCII” or other
form. Any alternate format must include the full Project
Gutenberg™ License as specified in paragraph 1.E.1.
• You pay a royalty fee of 20% of the gross profits you derive from
the use of Project Gutenberg™ works calculated using the
method you already use to calculate your applicable taxes. The
fee is owed to the owner of the Project Gutenberg™ trademark,
but he has agreed to donate royalties under this paragraph to
the Project Gutenberg Literary Archive Foundation. Royalty
payments must be paid within 60 days following each date on
which you prepare (or are legally required to prepare) your
periodic tax returns. Royalty payments should be clearly marked
as such and sent to the Project Gutenberg Literary Archive
Foundation at the address specified in Section 4, “Information
about donations to the Project Gutenberg Literary Archive
Foundation.”
• You comply with all other terms of this agreement for free
distribution of Project Gutenberg™ works.
1.F.
Most people start at our website which has the main PG search
facility: www.gutenberg.org.